To describe our current procedure of robot-assisted laparoscopic radical prostatectomy (RALP), and to assess the effect of the learning curve on perioperative data, early oncological outcomes and functional results, as RALP has increasingly become a treatment option for men with localized prostate cancer. In all, 206 consecutive men had a RALP between July 2001 and November 2008 for localized prostate cancer. Among the overall cohort, the 175 men operated on by the same surgeon were distributed into five groups according to the chronological order of the procedures. The mean follow-up after RALP was 18.3 months. Patient demographics, surgical data and postoperative variables were collected into a prospective database. Data were compared by chronological groups into single-surgeon cohort. The median operative time and blood loss were 140 min and 350 mL, respectively. The complication rate was 8.3%. Cancers were pT3-4 in 34.5%. The mean hospital stay and duration of bladder catheterization were 4.3 and 8.2 days, respectively. The rate of positive surgical margins (PSMs) was 17.2% in pT2 cancers. The recovery rate of continence was 98% at 12 months. Intraoperative time, blood loss and length of hospital stay were significantly improved after a short learning curve. The continence recovery, the rate and the length of PSM were also improved beyond the learning curve, but difference was not statistically significant. RALP is a safe and reproducible procedure and offers a short learning curve for experienced laparoscopic surgeons. Beyond the learning curve, continued experience might also provide further improvements in terms of operative, pathological and functional results.